individual n se

8/6/2019 Individual n SE

1/42

IZA DP No. 3974

Why Do Individuals Choose Self-Employment?

Christopher Dawson

Andrew HenleyPaul Latreille

D

IS

C

U

S

S

IO

N

P

A

P

E

R

S

E

R

IE

S

January 2009


2/42

Why Do Individuals ChooseSelf-Employment?

Christopher Dawson

Swansea University

Andrew HenleySwansea University

and IZA

Paul LatreilleSwansea University

Discussion Paper No. 3974January 2009

IZA

P.O. Box 724053072 Bonn

Germany

Phone: +49-228-3894-0Fax: +49-228-3894-180

E-mail: [email protected]
mailto:[email protected]:[email protected]


3/42

IZA Discussion Paper No. 3974January 2009

ABSTRACT

Why Do Individuals Choose Self-Employment?*

This paper undertakes an analysis of the motivating factors cited by the self-employed in theUK as reasons for choosing self-employment. Very limited previous research has addressedthe question of why individuals report that they have chosen self-employment. Two questionsare addressed using large scale labour force survey data for the UK. The first concerns the

extent to which the self-employed are self-employed out of necessity, opportunity, lifestyledecision or occupational choice. The second concerns the extent to which there isheterogeneity amongst the self-employed on the basis of the motivations that they report forchoosing self-employment. Factor analysis reveals a number of different dimensions ofentrepreneurship on the basis of stated motivation, but with no evidence that being forcedinto entrepreneurship through economic necessity is a significant factor. Motivation towardsentrepreneurship is therefore highly multidimensional. Multivariate regression analysis isemployed using a method to control for self-selection into self-employment. This reveals

significant differences between men and women, with women concerned more with lifestylefactors and less with financial gain. Market-directed opportunity entrepreneurship is morestrongly associated with higher educational attainment. Those joining family businessesappear not to value prior educational attainment. Public policy to promote entrepreneurshiptherefore needs to be tailored carefully to different groups.

JEL Classification: L26, J24

Keywords: self-employment, entrepreneurship, motivation, occupational choice

Corresponding author:

Andrew Henley


4/42

1. Introduction

Self-employment is a form of labour market status which may encompass a wide

range of different activity. Individuals may choose to be self-employed for many different

reasons, and as a result the self-employed as a group may be highly heterogeneous. At one

end of a possible spectrum the self-employed may be identified as entrepreneurial, single

employee micro-businesses. A substantial body of research investigates the self-employed as

entrepreneurs, using self-employment as an observable category which, albeit imperfectly,

identifies the stock of entrepreneurial talent in the economy. At the other end this spectrum,

self-employment may comprise a far less desirable state chosen reluctantly by individuals

unable to find appropriate paid employment under current labour market conditions. So, for

example, individuals wanting flexible working hours might choose self-employment if a paid

employment contract offering sufficient flexibility is unavailable. For some self-employment

may be chosen as the only available alternative to unemployment. Indeed in many developing

economies self-employment may be viewed as a form of informal sector employment

activity.

To gain an understanding of the positive reasons why individuals might choose to

become self-employed in order to venture a new business, a range of underlying motivations


5/42

reported levels of interest in entrepreneurship are often very high, particularly if questions are

framed loosely in terms of open-ended career aspirations (see Blanchflower et al, 2001). Such

high levels are difficult to square with much lower levels of actual self-employment in many

countries. This suggests that intentions to become self-employed are either frustrated or that

self-reported aspirations are vague and poorly formed. A further literature in economics

models the choice of self-employment as an occupational choice decision, focusing on the

importance of background and demographic associations. A key objective of this literature,

taking its lead from the seminal study of Rees and Shah (1986), is to address whether the

choice of self-employment might be motivated by the expected income differential between

self-employment and employment for a given individual (see Le, 1999 and Parker, 2004, for

surveys).

Very little, if any, research has addressed the question of why those who have actually

chosen self-employment, made that choice. The preceding discussion suggested that the self-

employed can be classified into two broad types; those that have entered voluntarily for

reasons such as independence, job satisfaction and or anticipated higher incomes, and those

that have been pushed into self-employed because of the absence of any other attractive

alternative. If the motivations behind the decision to become self-employed are largely the

former then self-employment can be viewed positively, providing the opportunity for


6/42

employment may be short and disheartening, and in some cases may simply result in delayed

transition into unemployment.

In this paper we undertake an analysis of the motivating factors cited by the self-

employed in the UK as reasons for choosing self-employment over the alternative of paid

employment working for someone else. This is undertaken by analyzing information

contained within selected years of the UK Labour Force Survey, the main quarterly British

household survey of labour market activity. The principal contribution of the paper is to

provide a multivariate analysis of the association of a range of demographic and background

characteristics with differing motivations for choosing self-employment, whilst controlling

for the selection bias that arises because the self-employed as a group are unlikely to be

representative of the whole population of the economically active. Specifically then we are

concerned with two questions. The first concerns the extent to which the self-employed are

self-employed out of necessity, opportunity, lifestyle decision or occupational choice. The

second concerns the extent to which there is heterogeneity amongst the self-employed on the

basis of the motivations that they report for choosing self-employment. A factor analysis

finds the existence of a number of different dimensions of entrepreneurship on the basis of

stated motivation. We conclude that there are significant differences between different types

within the self-employed in terms of gender, ethnicity, educational attainment and housing


7/42

2. Background and Previous Literature

Entrepreneurship, as a topic for research, has grown dramatically in recent years

primarily due to the relationship thought to exist between entrepreneurial activity and

economic development. New firms are thought to create new employment opportunities

(Parker and Johnson 1996, Ashcroft and Love 1996). New firms are also thought to be

involved significantly in innovative activity, such that the role of innovative entrepreneurship

is viewed as a key transmission mechanism between the creation of knowledge and economic

growth (Audretsch, 2007). In addition, self-employment is an important occupational option

for many in the labour force. At any one time it may account for approximately a tenth of all

employed workers (Evans and Leighton, 1989).1 The perceived economic importance of

entrepreneurial activity has thus spawned extensive research attempting to understand the

characteristics of potential entrepreneurs, and the process of transition into entrepreneurship.

Whilst much previous empirical work has sought to determine what personal

characteristics separate the self-employed from the employed, less attention has been given to

the reasons that individuals cite for choosing self-employment. Moreover, very few previous

studies employ nationally representative data, and those that do, have generally refrained

from the further analysis of background characteristics that may be associated with these


8/42

contrast, necessity entrepreneurs are pushed into self-employment because of negative

external forces, such as layoff and a subsequent lack of available paid-employment work.

At the aggregate level research has addressed the case for both push and pull

theories by examining the relationship between self-employment and unemployment. There

is little agreement here. Theoretical arguments have been constructed in support of both a

positive and a negative relationship between self-employment and unemployment. On one

hand, the prosperity-pull hypothesis suggests that individuals are more likely to attempt to

start a new firm under conditions of economic expansion, when incomes are growing and

opportunities are strong for market specialisation. Accordingly, a higher new firm formation

rate may be associated with lower local unemployment. High unemployment will inhibit the

market demand for products of the self-employed, and expose those who are self-

employment to greater risk of falling incomes and possibly bankruptcy. This implies a

negative relationship between self-employment and unemployment.

On the other hand, according to the push hypothesis, increasing levels of

unemployment reduce the prospects for finding paid employment; as a result the expected

returns from entrepreneurship become more attractive, pushing people into self-employment

(Storey, 1982; Storey and Johnson, 1987). Moreover, second hand capital becomes both

cheaper and more readily obtainable as business closures increase in a time of recession


9/42


10/42

very important motivating factor. 54 per cent reported that both greater control over their

life and building something for the family were very important motivating factors for

forming their own business. In a very small-scale study, Hughes (2003) uses data from 61

Canadian female respondents in the province of Alberta who are self-employed. In general

push factors were not found to be the primary motivator behind the decision to become self-

employed. In contrast most were motivated by reasons such as independence and a positive

working environment.

Similarly, using data from a much larger nationally representative survey of 3,840

self-employed Canadians for 2000, the same author reports that independence/freedom is

the most important motivator for both men and women when entering into self-employment

(Hughes, 2006). 42 per cent of men and 24 per cent of women cite this as their main

motivation. For women, work-family balance and flexibility of hours were the next most

cited reasons. For men challenge and prospects of more money were the next most cited

motivators. Overall Hughes suggests that over 71 per cent of men can be classified as

opportunity entrepreneurs, with only 22 per cent pushed into self-employment as a result of

the lack of other suitable opportunities. A further 7 per cent of men could be classified as

work-family entrepreneurs. For women the percentages were 53 per cent, 22 per cent and

25 per cent respectively, demonstrating for women the far greater importance of work-family


11/42

cent of men cited unemployment/redundancy as a motivating factor behind choosing self-

employment, whereas by 2000 this figure had risen to 26 per cent. For women the results

were 5 and 10 per cent respectively. The most cited reason for men in both 1986 and 2000, at

39 per cent, was a preference for being ones own boss. For women in 1986 the most cited

motivating factor was going into business with a family member. By 2000 women cited

going into business with a family member equally at 23 per cent with prefer to be own

boss and to follow an interest as the most popular reasons. These results suggest that for

both men and women, despite lower rates of UK unemployment in 2000, the self-employed

seemed to have become more pessimistic about the availability of paid employment. Carter

et al (2003), find evidence from US data that the motivations offered by nascent

entrepreneurs when starting a business are dominated by self-realisation, financial success,

innovation and independence.

Frey and Benz (2003) assess the level of job satisfaction of the self-employed

compared to those in organizational employment using individual-level longitudinal data for

Germany, Switzerland and the UK. Although these data sources are not specifically

concerned with explicit motivations for choosing self-employment, the authors are able to

conclude that higher levels of subsequently reported job satisfaction amongst those

transitioning into self-employment are associated with an absence of hierarchy and sense of


12/42

provides support for self-employment/paid-employment decisions being made rationally on

the basis of a consideration of the expected earnings differential in the two states. Taylor

(1996) also finds that individuals appear to be attracted to self-employment because of higher

expected earnings. Similarly, Clark and Drinkwater (2000) suggest that the difference

between an individuals predicted earnings in paid and self-employment exerts a powerful

influence upon the employment decision. On the other hand other studies, including Gill

(1988) and Earle and Sakova (2000), question this finding. These mixed empirical results

may be explained by the divergence of data sources used, the robustness of model

specifications and identifying exclusions, the ambiguous nature of the definition of

employment or self-employment and the difficulties in accurately measuring self-

employment income and therefore the differential between the two states. It is therefore open

to debate as to whether the expected financial gain from choosing self-employment is a

predominant factor in determining occupational status, or whether other non-pecuniary

considerations are of equal or greater importance.

Whilst this literature offers some pointers to potential motivations behind self-

employment as occupational choice, the direct survey evidence which is available is limited

and usually based upon small samples under non-random selection criteria. Moreover, whilst

an expansive literature exists on which personal characteristics separate entrepreneurs from


13/42

3. UK Labour Force Survey Data Source

The data used in the present study is obtained from the United Kingdom Quarterly

Labour Force Survey (QLFS), covering the years 1999-2001. The QLFS is a national UK

household survey in which all adults at each sampled postal address are asked about current

labour market status and activity. The survey is used by the British government to provide

important labour market intelligence data, but is also made available, after a certain time lag,

to the research community in anonymised form for other secondary analysis. Although, the

QLFS has a panel design, with each household of the sample interviewed for 5 consecutive

quarters, we are primarily interested in individual responses to a schedule of recall

questions about self-employment choice asked of each individual, and so include only one

observation on each individual for analysis.3

Interviews were achieved at approximately

59,000 addresses in each quarter, resulting in a sample of approximately 138,000 individual

adult respondents in each quarter. The very large size of the QLFS means that it is possible to

obtain a large sample of observations on the self-employed, facilitating robust analysis of

particular sub-groups. The QLFS questionnaire includes schedules of household and

individual questions covering family structure, housing information, economic activity,

employment, educational and health issues.


14/42

2. wanted more money

3. for better conditions of work

4. family commitments / wanted to work at home5. opportunity arose capital, space, equipment available

6. saw the demand / market

7. joined the family business

8. nature of the occupation

9. no jobs available (locally)

10. made redundant

11. other reasons

12. no reason given

(Source: QLFS questionnaire, Spring Quarter (March to May) 1999)

Each individual respondent was asked to choose up to four reasons. Individual observations

are pooled across the three available Spring quarters providing a total pooled sample of

147,686 economically active individuals, of which 17,507 (11.9 per cent) are self-employed.4

As a result of multiple choices there are 23,851 choice responses to the question for these

17,507 self-employed respondents. Table 1 reports the proportions of the self-employed who

provide each reason for becoming self-employed using the pooled data over the period 1999

to 2001. Table 2 reports the proportions of total responses for each reason given, providing

columns which sum to 100 percent.

Overall Tables 1 and 2 suggest that, for the majority of respondents, entry into self-

employment is influenced by a range of non-pecuniary factors, of which the need for

independence is the most pronounced. This finding supports the conclusions in Dennis

(1996) and in Frey and Benz (2004) who suggest the attraction of entrepreneurship is work


15/42

Other motivations, reflecting pull factors, attract significant response levels. 5.4 per

cent of respondents cite the desire for better working conditions and 8.8 per cent indicate as

a motivation that they saw the demand for the product or service which they felt they had to

potential to provide. The least cited reason for choosing self-employment is the negative

motivation of a lack of availability of jobs locally. The nature of this option directly focuses

attention on those respondents who were forced into self-employment as the only viable

alternative to unemployment. Only just over 3 per cent of individuals indicate that the lack of

jobs available locally was a motivation behind their transition into self-employment. This

implies that for most the decision to become self-employed is overwhelmingly a positive

action. However, whilst there appears to be a low proportion motivated by the availability of

jobs locally it is important to note that the UK unemployment rate averaged only around 6

per cent over the time period in question. The restrictive nature of this question makes it clear

that those individuals who cite this reason have been pushed into selecting self-employment.

Many of the other possible motivations, such as more money, joined the family business

and family commitment could operate as either push or pull factors. For example, more

money could be suggestive of attractive higher earnings in self-employment or could be

suggestive of perceived poor wages in the paid-employment sector. Moreover, motivations

that might typically be considered as attractors for example, to be independent and better

conditions of work, can undoubtedly also operate as push factors. That is, a lack of


16/42

We turn now the different pattern of responses between men and women. Table 1

reports the results of t-tests for the significance of the difference in the response rates

between men and women for each motivating reason. In all but two cases these differences

are statistically significant. However independence and nature of the occupation are the

two most commonly cited motivations for both men and women, although the proportion of

women citing independence is considerably lower than for men. The most important

difference in the pattern of responses is that nearly 22 per cent of women cite family

commitments as a reason for choosing self-employment, compared to only 2 per cent of

men. Corresponding to the importance of balancing family and work for women, it is also

clear that women are less likely to cite financial gain as a reason for choosing self-

employment. Self-employment for women is far more likely to be framed in broader quality

of life terms, than in terms of narrow pecuniary advantage.

As an exercise in attempting to identify key dimensions in the pattern of reasons given

for choosing self-employment, Table 3 reports the results of a factor analysis. Since it is clear

that key demographic factors may be correlated with the pattern of responses, the factor

analysis includes gender and age as well as the different reasons for choosing self-

employment. The method of estimation is maximum likelihood and the preferred number of

factors which results is seven. Table 3 reports the eigenvalues of each factor in the first row

and factor loadings for each variable (age female and eleven different reasons for choosing


17/42

1. idiosyncratic entrepreneurship; those choosing self-employment regardless ofopportunity and fit with professional background for other reasons;

2. market-facing opportunity entrepreneurship; those choosing self-employment becauseof perceived external opportunity, but contrary to prior professional background;

3. professional entrepreneurship; those choosing self-employment to join professionalpartnerships and establish a professional practice;

4. family entrepreneurship; those choosing self-employment to join family businesses,some of which may be professional in nature;

5. independence-seeking entrepreneurship; those actively attracted to self-employmentby a desire for independence, not associated with economic necessity;

6. lifestyle entrepreneurs; those, particularly women and younger people, who chooseself-employment because it offers improved ability to balance work and family

commitments, but not associated with economic necessity;

7. reward-seeking entrepreneurship; those younger people voluntarily leaving paidemployment to pursue actively financial reward and improved working conditions.

This exercise in data description points to significant heterogeneity in the motivations

behind particular individual decisions to choose self-employment. The majority of these

dimensions entail considerable elements of pull. In fact it is noticeable that push factors


18/42

entrepreneurship appear to be related to the type of entrepreneurial activity (professional,

family business), the need for independence and/or financial reward, and, particularly in the

case of women, lifestyle considerations.

4. Methodology

The purpose of this section is to describe the empirical methodology used to model

associations between a range of background and demographic factors and particular

motivations for choosing self-employment. This is in order to understand more about which

individuals are more likely to indicate particular motivations for self-employment. The

approach adopted recognizes that the self-employed are a non-random sample of the

economically active, and therefore that any modeling of the motivations of the self-employed

must control for the non-random nature of the sample. This is undertaken using a selection

bias correction method. Uncorrected estimates of the association between a certain

background factor and one of the motivations will typically provide spurious indicators of the

statistical significance of that association.

The conventional approach to this statistical issue has been to employ a Heckman

selection correction (Heckman 1979). However identification in the Heckman model rests on


19/42

empirical grounds be restricted to zero. In the present case it seems unlikely that such

restrictions might exist ex ante, since the underlying processes which determine whether

someone is self-employed are likely to very similar to those which determine the reason why

someone chose to be self-employed.

Sartori (2003) develops an estimator where selection and outcome equations can share

a common structure and where identification is achieved on the assumption of identical errors

in the selection and outcome equations. Equations (1) and (2) show selection and outcome

processes under the case of non-random selection for individual i from a sample 1 to n. The

dependent variable in equation (1) represents sample selection on the basis of an underlying

standardised continuous process. The dependent variable in the outcome equation (2) is also

unobserved, standardised and continuous. The explanatory variables,x, are the same in both

selection and outcome equation, with and denoting the different coefficient vectors in

each. Each equation contains a normally distributed, mean zero error term, v1 and v2.

(1)iii vxU 11 ' +=

(2)iii vxU 22 ' +=

Instead of observing the Us, we observe two dichotomous variables, and , shown as

the dependent variables in equations (3) and (4). The relationship between the unobserved

selection and outcome processes and the observed dichotomous variables is as follows:

iZ1 iZ2


20/42

into account the selection process. This can be described by defining three random variables

such that:ijY

.otherwise0and,1and1if1

;otherwise0and,0and1if1

;otherwise0and,01

212

211

10

===

===

==

iii

iii

ii

ZZY

ZZY

ZifY

(5)

iY0 has a value of 1 if the observation is not selected, has a value of 1 if the observation is

selected but the value of the outcome variable is 0, and has a value of 1 if the observation

is selected and the value of the outcome variable is 1. In order to construct a likelihood

function for the model it is necessary to specify the data generating process for the

probability that in each case. As shown by Sartori (2003) these probabilities are as

follows:

iY1

iY2

1=jiY

>==

=

==

==

0)(if)(

0)(if)()1Pr(

otherwise0

0)(if)()()1Pr(

)()1Pr(

2

1

0

ii

ii

i

iii

i

ii

xx

xxY

xxxY

xY

(6)

is the cumulative standard normal density function. The likelihood function is then

conventionally defined as the product of the different probabilities, Pji , for each combination


21/42

The critical assumption here is that the error terms in the selection and outcome

equations are identical. It is important to assess this model restriction. The assumption of

(near) identical errors can be expected to hold when the decision processes behind selection

and the outcome of interest are very similar, and result from the same causal process (see

Sartori, 2003, p. 112). In the present case this seems reasonable, since, as already noted, the

process governing the decision to be self-employed is likely to be very similar to that behind

the motivation for that decision. Sartori also notes that the two processes should occur at the

same time and/or in the same place, and in the present case this is by definition true.

For the purposes of the selection analysis an individual is defined as self-employed

(i.e. Z1i = 1) if they reported their employment status in their main occupation as self-

employment and gave at least one response to the question concerning their motive for

choosing to become self-employed. The non-selected group are the paid-employed and are

defined as those individuals reporting their employment status in their main occupation to be

an employee.

The choice of which covariates to include as having potential association with

different motivations for choosing self-employment is to some extent constrained by the

nature of the QLFS data source. The QLFS is a deliberately designed as a large survey in

order to allow the derivation of official estimates of labour market activity on a detailed


22/42

alongside other basic demographic information including age (in a non-linear quadratic form)

and marital status. Membership of an ethnic minority is also included, since the relationship

between ethnicity and entrepreneurial activity is one that has figured in previous literature.6

The potential role of household structure is captured through the inclusion of the number of

dependent children under the age of 16. Self-employed activities are higher heterogeneous

and it is therefore important to control for the role of education in order to assess the extent to

which both higher educated individuals may be motivated to choose self-employment as a

route to professional status and less educated individuals may be motivated to choose self-

employment due to a lack of other economic alternatives. Educational attainment is captured

in the model through a series of dichotomous highest level of educational attainment

variables. These are: university or college degree level; other non-degree higher education;

A-levels or equivalent (post-compulsory examinations taken at 18 as qualifying exams for

college or university entrance), GCSE or O-levels (age 16 schooling attainment

qualifications); and other qualifications. The literature on education and self-employment is

mixed; arguments can be made for both a positive and a negative relationship. Skills

associated with successful entrepreneurship may not necessarily be those skills obtained from

formal qualifications. However, those with higher levels of education may select themselves

into professional occupations where self-employment status, perhaps within the context of a

professional partnership is more common. Housing tenure status is also included, not least


23/42

is pooled over three years, year dummy variables are included to capture any effect on stated

motivations of changing aggregate economic or societal conditions.

5. Empirical Results

Table 4 reports estimated marginal effects from the regression of the selection

equation. Only one equation is reported since the marginal effects of the selection equation

remain identical for each different outcome, with only very minor variation in levels of

significance. Levels of statistical significance are very high for most of the covariates. Older

individuals are, other things equal, more likely to be self-employed, but the significance of

the quadratic term suggests that the likelihood of self-employment increases at a declining

rate. Women are, other things equal, four percentage points less likely to be self-employed, a

result that is well-established in the literature. The disabled are 0.6 percentage points more

likely to be self-employed, almost certainly reflecting the greater flexibility in working

conditions that self-employment may offer. Members of ethnic minorities are nine percentage

points more likely to be self-employed. The presence of dependent children raises the

probability of self-employment. The precise explanation for this association is uncertain, but

it may be related again to greater working flexibility. There is no statistically significant

association between likelihood of self-employment and marital status. However the signs on

the coefficients are plausible; negative for the married and positive for the widowed or


24/42

employment. All tenure groups are more likely, other things equal, to be in self-employment

compared to those in social rental housing. Outright owners are, other things equal, over 4

percentage points more likely to be self-employed. This may reflect access to capital

resources which can be used to provide collateral for business finance. Finally the year

dummies show that the rate of self-employment falls very slightly in the later years.

Table 5 reports the estimated marginal effects for each outcome equation. Results will

be discussed for each group of covariates in turn across the different motivations for choosing

self-employment. Turning first to age, it is clear that the strongest positive association is with

independence older individuals appear to be more likely to value independence as a

positive attribute associated with entrepreneurship, and the results suggest a positive but

decreasing association with age up to 56 years of age.. All other motivations are positively

associated with age, but at a declining rate as indicated by negative quadratic terms; however

the sizes of the associations are much smaller than for independence.

There is a negative association between being female and the likelihood of stating a

particular motivation for self-employment in all cases except for family / home. These

negative coefficients reflect lower levels of self-employment generally amongst women.

However it is clear that women are significantly more likely than men to choose self-


25/42

their male counterparts and are less likely to choose occupations more commonly associated

with self-employment.

Disability is in most cases positively associated with the different motivations for

choosing self-employment, reflecting the positive association between self-employment

status and being disabled. The strongest associations are for working conditions,

occupation, no jobs and in particular other reasons. It may be the case that the disabled

are more likely to be pushed rather than pulled towards self-employment, either because of

segregation into particular occupations or because of discrimination in the paid employment

market. To this extent it is noticeable that the more economic motivations such as money

have smaller coefficients, and that the only negative (albeit insignificant) coefficient is in the

equation for opportunity arose.

Membership of an ethnic minority is in most cases positively associated with the

different reasons for choosing self-employment. Ethnic minorities are particularly likely,

other things equal, to state independence as a reason (marginal effect: nine percentage

points). Other significant reasons are family business and no jobs. These findings are

consistent with the view that members of ethnic minorities may choose self-employment

because they are excluded from the formal labour market, and may prefer to build business


26/42

dependent child raises the likelihood of stating money as a motivation by 2.6 percentage

points. On the other hand having a dependent child also raises the attractiveness of self-

employment for those seeking independence (marginal effect of 4.2 percentage points per

child) and flexibility to deal with family and home circumstances (marginal effect of 3.1

percentage points per child).

The marital status controls attract a mixed set of coefficients. Both those who are

married (or cohabiting) and who were formerly married (widowed, divorced or separated) are

more likely to report money as a motivation compared to the never married. In either case

however the motivations may be different. In the case of the formerly married, bereavement

or separation may have resulted in financial distress and the need to increase income. For the

married the motivation may be more aspirational related to a stronger desire to build a

home compared to those not married. Thus the results show that, other things equal, those

who are married are 2.4 percentage points more likely to cite family / home as a motivation,

reinforcing the association already noted above with dependent children The results show

negative associations between being married and independence (marginal effect of -3.3

percentage points) and occupation (marginal effect of -2.5 percentage points). Being

married may be associated with a stronger desire for financial security which is in turn

associated with a greater tolerance for building a career working for an organization. Those


27/42

The results in Table 5 show a complex and varied association between educational

attainment and different motivations for choosing self-employment. The better educated, in

particular those individuals with degree-level qualifications, are more likely to report that

self-employment was chosen for such reasons as independence, better working conditions

and the nature of their occupation. In particular, the table shows significant positive

marginal effects for university graduates compared to someone with no formal qualifications

for independence (3.2 percentage points), for working conditions (1.5 percentage points)

and for occupation (7.4 percentage points). Degree holders are also very significantly more

likely to cite other reasons (6.2 percentage points). More educated individuals are also less

likely to report family business, no jobs or redundancy as motivations. Graduates in

particular are significantly more likely than others to report that self-employment was chosen

because of the nature of their occupation, consistent with self-employment being a normal

occupational status for sizeable numbers of university-educated professionals. It was noted in

Table 4 that those whose highest academic qualification is A-levels (post-compulsory school

examinations at age 18) are significantly more likely to be in self-employment. The pattern of

coefficients in Table 5 for this group is rather different, with a number of significant positive

associations with various motivations, especially independence (marginal effect of 10.8

percentage points) and money (marginal effect of 5.4 percentage point). In effect this group

comprises individuals who have failed to achieve a standard of academic achievement for


28/42

Those individuals who have no formal qualifications are more likely to enter into self-

employment for reasons such as no jobs available locally, joined the family business and

more money compared to most skill levels. These results are consistent with the unskilled

being more likely to encounter greater problems in finding work, and less likely subsequently

to command a high wage. Therefore individuals with no formal qualifications are

significantly more likely to enter into self-employment due to the both the lack of jobs

available locally and for pecuniary reasons, and therefore may ultimately find self-

employment to be a less fulfilling occupational choice. Similarly, those individuals who

report family business as a motivation are less likely to have acquired skills and more likely

to have left education at an early stage in order to start working for that family business.

Entry into self-employment may for such individuals have presented itself as a straight-

forward family expectation. Others with educational qualifications are significantly less

likely to have entered self-employment because of a family business, with marginal

probability effects ranging from around minus two percentage points for those with age 16

school qualifications to minus nine percentage points for higher education graduates.

As noted above home-ownership and private sector renter status are significantly

positively associated with self-employment. Consistent with this finding in Table 4, Table 5

reveals a wide range of significant associations between home ownership and different

motivations for choosing self employment For every motivation excluding no jobs available


29/42

explain why those renting social housing are almost always less likely to report a particular

motivation. Social housing status is associated with a wide range of other factors associated

with poverty, such as low income, low skill and single parent status. These work additively to

contribute to lack of economic opportunity. This self-employment seems much more likely to

associated with forced entrepreneurship. However, as the discussion above has highlighted,

this type of self-employment appears to comprise only a small proportion of the total and

does not figure as a significant dimension of entrepreneurship.

5. Discussion and Conclusions

The existing literature providing large scale survey evidence on why individuals

choose self-employment is very limited. We currently know a good deal about who chooses

self-employment, but not very much about why. This paper has analyzed data which is

available over a particular time period between 1999 and 2001 within the UK Quarterly

Labour Force Survey, a very large survey which asks a significant sample of the self-

employed to indicate up to four choices from a list of eleven possible reasons for their

decision. Clearly some of these respondents will have been self-employed for some

considerable length of time, while others may have only recently transitioned from another

economic status. To that extent responses may be subject to some unknown element of recall

bias, or ex post rationalization. Nevertheless such data is the best that is typically available.


30/42

employed as a group. In very broad terms the different motivational dimensions can be

summarized as relating to the existence of opportunity to start of business, the nature of an

individuals profession (if they have one), the desire for a particular lifestyle and need to

balance family commitment with working life, and finally the opportunity afforded by having

resources available to support a new business venture.

Significant differences in the pattern of response are apparent for certain groups.

Women are much more likely to report lifestyle and family reasons for choosing self-

employment than men a conclusion that is perhaps not surprising but does imply issues of

equal opportunity. Women are less likely to report financial gain as a motivating factor. Two

other groups for whom lifestyle issues figure more importantly are older individuals and

members of ethnic minorities. However for the latter group, care must be taken in the

interpretation of this finding, since we cannot rule out the possibility that it may be culturally

more acceptable to provide a justification other than one which is financial for certain groups.

What we have termed opportunity entrepreneurship appears from the results to be

associated with educational attainment. More educated individuals appear to be more likely

to view self-employment in positive terms, offering independence and financial reward, as

well as better working conditions. The least educated individuals, that is those lacking any


31/42

poorly because of a lack of appreciation of the value of education and skills. This finding

accords with other recent research on the relationship between management practice and firm

performance (Van Reenen and Bloom, 2007). This research concludes that family-run

businesses are significantly less likely to employ good management practices.

The present study has found little direct evidence for forced entrepreneurship; that is

significant numbers of individuals who appear to have chosen self-employment out of

necessity because of loss of previous paid employment and a lack of other paid alternatives.

The vast majority choose self-employment for positive reasons. However it should be noted

that the time period covered by the data analyzed extends across the middle of a period of

sustained economic growth in the UK economy. Whether this conclusion would be as robust

during the current period of severe economic downturn and rapidly rising unemployment is

open to debate and, with suitable data, further future analysis. Indeed it would give cause for

considerable concern if the proportion choosing self-employment because it represents the

only alternative to economic inactivity was to rise significantly in the next few years. Such

forced choices may not lay solid foundations for well-resourced, successful new business

ventures.

For public policy, the finding of significant heterogeneity amongst reasons for


32/42

observers may be left pondering how many of these individuals actually achieve the

significant improvement in earnings that they appear to desire. This in turn points to the need

for policy interventions designed to promote more careful business planning amongst

aspiring entrepreneurs. By contrast what might be termed supply-side factors appear to play a

relatively larger role opportunity entrepreneurs frame opportunity in terms of possessing

sufficient resources to take advantage of a potential opportunity, rather than in terms of

available market. Finally a small proportion of the self-employed chose this state because

they joined a family business. Such individuals do not appear to be particularly well educated

and may fail to appreciate the contribution that academic qualifications and other skills may

make to business performance. Policy intervention needs to be designed carefully to target

such individuals.


33/42

Footnotes:

1Within the UK, self-employed individuals account for approximately 12.5% of the work-

force. Blanchflower (2000) provides recent comparative data for OECD economies.

2Previous literature has looked at the gender differences between motivations to become

self-employed, and identified the importance of work-family balance issues for women. See

Buttner and Moore (1997) and Hughes (2006) for an overview of the evidence.

3The choice of time period 1999 to 2001 is constrained by the availability of the question on

motivation for becoming self-employed. This question has been asked only periodically in

the QLFS and was dropped after 2001.

4Each member of the QLFS sample is interviewed for five consecutive quarters in order to

provide a rotating longitudinal element to the survey. This means that the spring quarter files

for 2000 and for 2001 included two observations on those who were self-employed in each

year, and therefore a duplicate (although potentially inconsistent) response to the question on

reasons for becoming self-employed. To avoid duplicate observations in our analysis, those

individuals in the spring quarter 2000 sample who were also included in the spring quarter

1999 sample, and those in the spring quarter 2001 sample who were also included in the

spring quarter 2000 sample, were deleted from the analysis. In principle one could have

deleted the first rather than the second duplicate observation. Both methods were

investigated, and it was found that the results of the secondary analysis in each case were


34/42

References

Ashcroft, B.K. and Love, J.H. (1996), Firm births and employment change in Britishcounties, Papers in Regional Science, 75(4): 1-18.

Audretsch, D.B. (2007), Entrepreneurship capital and economic growth, Oxford Review of

Economic Policy, 23(1): 6378.

Black, J., de Meza, D. and Jeffreys, D. (1996), House prices, the supply of collateral and the

enterprise economy,Economic Journal , 106(434): 60-75.

Blanchflower, D.G. (2000), Self-employment in OECD countries, Labour Economics, 7(5):

471-505.

Blanchflower, D.G. and Meyer, B.D. (1994), A longitudinal analysis of the young self-

employed in Australia and the United States, Small Business Economics, 6(1): 1-19.

Blanchflower, D.G., Oswald, A., and Stutzer, A. (2001), Latent entrepreneurship across

nations.European Economic Review, 45: 680-691.

Buttner, E.H. and. Moore, D.P (1997), Womens organizational exodus to entrepreneurship:

self-reported motivations and correlates with success, Journal of Small Business

Management, 35(1): 34-46.

Carter, N.M., Gartner, W.B., Shaver, K.G. and Gatewood, E.J. (2003), The career reasons of

nascent entrepreneurs,Journal of Business Venturing, 18(1): 13-39.

Clark, K. and Drinkwater, S. (2000), Pushed out or pulled in? Self-employment amongt

Britains ethnic minorities,Labour Economics, 7: 603-628.

Cowling, M. and Mitchell, P. (1997), The evolution of UK self-employment: a study of

government policy and the role of the macroeconomy, The Manchester School, 65(4):

427-442.

Dennis, J. (1996), Self-employment: When nothing else is available?, Journal of LabourResearch, 17(4): 645-661.

Earle, J.S. and Sakova, Z. (2000), Business start-ups or disguised unemployment? Evidence

on the character of self-employment from transition economies, Labour Economics,


35/42

Hamilton, R.T. (1989), Unemployment and business formation rates: reconciling time series

and cross-section evidence,Environment and Planning, 21(2): 249-255.

Heckman, J. (1979),Sample selection bias as a specification error,Econometrica, 47(1): 153-

161.

Henley, A. (2004), Self-employment status: the role of state dependence and initial

conditions, Small Business Economics 22(1): 67-82.

Hughes, K (2003), Pushed or pulled? Womens entry into self-employment and small

business ownership, Gender, Work and Organization, 10(4): 433-454.

Hughes, K (2006), Exploring motivation and success among Canadian women entrepreneurs,

Journal of Small Business and Entrepreneurship,19(2): 107-120.

Krueger, N.F., Reilly, M.D., and Carsrud, A.L. (2000), Competing models of entrepreneurial

intentions.Journal of Business Venturing15: 411-432.

Le, A.T (1999), Empirical studies of self-employment, Journal of Economic Surveys, 13(4):381-416.

Parker, S.C. (2004), The Economics of Self-Employment and Entrepreneurship, Cambridge

(UK): Cambridge University Press.

Parker, S.C. and Johnson, P. (1996), Spatial variations in the determinants and effects of firm

births and deaths,Regional Studies, 30(7): 679-688.

Rees, H. and Shah, A. (1986), An empirical analysis of self- employment in the UK,Journal

of Applied Econometrics, 1(1): 95-108.

Robson, M.T. (1998), The rise in self-employment amongst UK males, Small Business

Economics, 10(3); 199-212.

Sartori, A (2003), An estimator for some binary-outcome selection models without exclusion

restrictions, Political Analysis, 11(2): 111-138.

Smeaton, D (2003), Self-employed workers: calling the shots or hesitant independents? A

consideration of the trends,Work, Employment & Society, 17(2): 379-391.

Storey D J (1982) Entrepreneurship and the New Firm Westport CT: Praeger Publishers


36/42

Table 1: Reported reasons for becoming self-employed - by individual

Reason (percentage) All Men Women

t-test

(p-value)

To be independent / a change 30.2 32.3 24.9 0.000

Wanted more money 12.7 14.5 8.0 0.000

For better conditions of working 5.4 6.0 4.0 0.000

Family commitments / wanted to work

at home 7.7 2.2 21.5 0.000

Opportunity arose - Capital, space,

equipment available 12.5 12.7 12.1 0.291

Saw the demand / market 8.8 8.8 8.6 0.722

Joined the family business 6.9 6.6 7.6 0.025

Nature of the occupation 21.5 21.1 22.5 0.030

No jobs available (locally) 3.4 3.7 2.4 0.000

Made redundant 9.3 11.6 3.6 0.000

Other reasons 14.6 13.8 16.6 0.000

No reason given 3.4 3.7 2.5 0.000

N 17507 12582 4925

Source: authors tabulations from LFS Spring Quarters 1999-2001

Notes: Columns do not sum to 100 per cent because respondents can give up to four reasons.

The right hand side reports the significance of a t-test for the difference between men and


37/42

Table 2: Reported reasons for becoming self-employed all responses

Reason (percentage) All Men Women

t-test

(p-value)

To be independent / a change 22.2 23.6 18.5 0.000

Wanted more money 9.33 10.6 6.0 0.000

For better conditions of working 4.0 4.4 3.0 0.000

Family commitments / wanted to work

at home 5.6 1.6 16.0 0.000

Opportunity arose - Capital, space,

equipment available 9.2 9.3 9.0 0.520

Saw the demand / market 6.4 6.4 6.4 0.978

Joined the family business 5.0 4.8 5.6 0.012

Nature of the occupation 15.8 15.4 16.8 0.009

No jobs available (locally) 2.5 2.7 1.8 0.000

Made redundant 6.8 8.4 2.7 0.000

Other reasons 10.7 10.1 12.3 0.000

No reason given 2.5 2.7 1.9 0.000

N 23851 17227 6624

Source: authors tabulations from LFS Spring Quarters 1999-2001

Notes: The right hand side reports the significance of a t-test for the difference between men

and women.Bold italic indicates p-value < 0.05.


38/42

Factor 1 Factor 2 Factor 3 Factor 4 Factor 5 Factor 6 Factor 7

Eigenvalue 1.158 1.060 1.105 1.142 0.853 0.724 0.578

Factor loadings:

Age 0.074 0.047 -0.040 0.049 -0.035 -0.139 -0.182

Female 0.029 0.002 0.003 0.051 -0.049 0.560 0.011

To be independent / a

change

-0.175 -0.071 -0.210 -0.510 0.813 0.000 -0.000

Wanted more money -0.081 -0.021 -0.093 0.187 0.009 -0.147 0.630For better conditions of

working

-0.052 -0.010 -0.041 -0.118 0.064 -0.023 0.234

Family commitments /

wanted to work at home

-0.056 -0.048 -0.065 -0.091 -0.139 0.558 0.059

Opportunity arose -

Capital, space,

equipment available

-0.237 0.962 0.095 0.089 0.037 0.000 0.000

Saw the demand /

market

-0.094 0.091 -0.006 -0.083 0.013 -0.007 0.066

Joined the family

business

-0.112 -0.138 -0.523 0.820 0.150 -0.000 0.000

Nature of the occupation -0.205 -0.288 0.865 0.330 0.134 -0.000 0.000

No jobs available

(locally)

-0.027 -0.029 -0.039 -0.069 -0.156 -0.043 -0.061

Made redundant -0.079 -0.040 -0.100 -0.129 -0.313 -0.234 -0.285Other reasons 0.991 0.110 0.050 0.047 0.035 -0.000 0.000

35

Table 3: Factor analysis of reasons for choosing self-employment

Source: authors computations from QLFS 1999-2001

Note: maximum likelihood method (LogL=-240.8); loadings >0.1 in italic; >0.2 inbold italic.

.


39/42

Table 4: Sartori estimates of self-employment selection equation

MarginalEffect p-value

Demographic factors:

Age .0041 .000

Age squared/100 -.0024 .000

Female -.0407 .000

Disabled .0058 .000

Ethnic minority .0088 .000Household and family status:

No. dependent Children


40/42

Table 5: Sartori estimates of motivation for choosing self-employment equations

(1)

Independence

(2)

Money

(3)

Working conditions

(4)

Family/home

(5)

OpportunityMarginal

Effect

p-

value

Marginal

Effect

p-

value

Marginal

Effect

p-

value

Marginal

Effect

p-

value

Marginal

Effect

p-

value


Age .0418 .000 .0061 .000 .0044 .000 .0063 .000 .0139 .000

Age squared/100 -.0373 .000 -.0060 .001 -.0043 .001 -.0051 .000 -.0104 .000

Female -.2853 .000 -.1557 .000 -.0582 .000 .0824 .000 -.0896 .000

Disabled .0066 .601 .0094 .240 .0152 .003 .0090 .034 -.0076 .353

Ethnic minority .0933 .000 -.0038 .758 .0132 .091 .0033 .626 .0017 .897Household and family status:



41/42

Table 5 (continued)

(6)

Saw the demand

(7)

Family business

(8)

Occupation

(9)

No jobs

(10)

RedundancyMarginal

Effect

p-

value

Marginal

Effect

p-

value

Marginal

Effect

p-

value

Marginal

Effect

p-

value

Marginal

Effect

p-

value


Age .00878 .000 .0021 .014 .0115 .000 .0039 .000 .0113 .000

Age squared/100 -.0070 .000 -.0004 .630 -.0061 .009 -.0031 .000 -.0087 .000

Female -.0627 .000 -.0279 .000 -.1544 .000 -.0314 .000 -.0883 .000

Disabled .0013 .856 -.0019 .682 .0256 .022 .0174 .000 .0042 .271

Ethnic minority .0217 .038 .0363 .000 -.0176 .322 .0258 .000 -.0247 .002Household and family status:



42/42

Table 5 (continued)

(11)

Other reasonSource: authors computations from QLFS 1999-2001Marginal

Effect

Notes: Models also include 12 regional controls coefficients not

reported. Full results available on request. Italic indicates p-value < 0.10,

bold italic indicates p-value < 0.05.

p-

value


Age .0049 .001

Age squared/100 .0020 .203

Female -.0704 .000

Disabled .0441 .000

Ethnic minority .0202 .134Household and family status:


individual n se

Documents